Evaluation of Fever Occurring in Labor in Patients Receiving Epidural Anesthesia

Historically the diagnosis of chorioamnionitis (an infection of the membranes surrounding the fetus) for patients in labor has been made on the basis of multiple clinical variables such as maternal fever, fetal tachycardia, uterine tenderness, or foul smelling vaginal discharge. The diagnosis also takes into account a clinical picture consistent with risk factors such as prolonged labor and prolonged rupture of membranes. Since most of these findings are not specific, chorioamnionitis becomes the diagnosis of exclusion unless there is another explanation for the fever. Randomized studies have clearly shown that maternal antibiotics and neonatal septic work-ups are indicated once the diagnosis of chorioamnionitis has been made. Neonatal sepsis is a severe infection of the blood stream. Such policies have important implications for health care providers, on the impact of medical costs, and on the duration of hospital stay. This becomes especially true for the newborn that is transferred to the neonatal intensive care unit (NICU) for sepsis evaluation. Often times, these newborns are prophylactically treated with antibiotics based on the suspicion of an infection, while waiting for finalized blood culture results. Since newborn sepsis is such a difficult diagnosis to make, many more newborns are treated than actually have the disease and the length of their hospital stay may be significantly increased.

Over the past 15 years, both observational and randomized trials have observed an increase in maternal fever associated with epidural anesthesia in labor. These trials have shown increased ranges from 10 - 15% over baseline rates and an increased relative risk of 1.5 to 15 fold, and even up to 70 fold in one study, over the rates seen in women not receiving epidural anesthesia. After correcting for duration of labor and other confounding variables, these increases remained present. Since epidural fever is virtually impossible to distinguish from chorioamnionitis-related fever, these women are almost all treated with antibiotics and given the diagnosis of infection. This approach also has tremendous impact on the evaluation and care of the newborn. Two specific studies evaluated this impact on the neonate. Lieberman found that babies of mothers given epidural anesthesia were more likely to be evaluated for sepsis (34 vs. 9%) and treated with antibiotics (15 vs. 4%). Similarly, Philips found the same increase (25 vs. 16% and 19 vs. 11%). Both studies had very low rates of confirmed neonatal sepsis. On a national basis, the cost of this confusion nationally is tremendous.

There is one known way to distinguish between true chorioamnionitis in labor and non-infectious fever due to the epidural anesthesia. Gibbs and colleagues found that amniotic fluid aspirated from an intrauterine pressure catheter, a device commonly utilized for monitoring contractions in labor, could be used to accurately make the diagnosis of infection using gram stain and culture. More recently, many papers have shown that low glucose levels and elevated IL-6 in amniotic fluid were also accurate tests for infection. While these markers could in theory be used for distinguishing between epidural fever and true chorioamnionitis, most patients do not require such a device and this approach would not likely gain widespread favor. Alternatively, however, this approach could be used as a research tool in women who already have such a catheter in place to determine if there are additional non-invasive clinical or laboratory markers to distinguish one from the other.

Recently proteomic assessment has become an extremely effective tool in determining if there are certain markers for various diseases. Proteomics is the determination of the structure, function, and expression of all of the corresponding proteins that are encoded within the genome structure. It can also be defined as the "fingerprint" of a disease process. It involves running tandem mass spectometry on the fluid of interest. Such an approach could be extremely valuable both in determining whether the mother actually has chorioamnionitis and, if so, whether there are better markers for neonatal sepsis.